How to look for content clues in your analytics

Our guest blogger this month is the wonderful Clare O’Brien, who heads up London-based content strategy and digital communications consultancy, CDA. She is also a woman on a mission to go beyond pure numbers to find insights that businesses can use to make content decisions…

Businesses have become fixated by metrics

Clare%20O%27BrienWe seem to be getting a bit confused between analytics and analysis. Broadly speaking – and thanks largely to the ubiquity and ease of access to Google Analytics (GA) – businesses have become fixated by traffic volumes, bounces, sources, journeys and subsequent destinations and the like and aren’t looking to learn more.

Those average and aggregate numbers are a) easy to come by and b) look pretty impressive in chart form for the boss. This means you don’t have to do too much more thinking. It also means missing some important clues in terms of discovering what content works and what doesn’t – and why.

As I wrote for Johnny Holland recently, anyone can sit beside a busy road, counting cars and compiling acres of data that can be easily turned into information about the makes and colours, directions travelled, volume densities and so forth.

Hardly anyone can use that same processed data to deliver back insights into journey purpose or satisfaction – this kind of knowledge requires more than counting. It needs us to ask people real questions, listen to their answers, understand their concerns and behaviours and use that knowledge to improve their road-using experiences. And so back to content.

Going beyond the numbers

As a profession, content strategists need to get much closer to audiences before, during and after we’ve engaged (or dis-engaged) them from the content that costs so much to produce and maintain. But it’s pretty hard, not to mention expensive, and takes a whole other client conversation and research workstream (read budget). Can we get smarter about understanding what content works and doesn’t for our audiences by looking more deeply into GA?

Well, yes, we can. We can look again at the data and information readily accessible from GA. It’s really about how we think about the data that’s already there.

For instance, keyword search logs and analysis. No magic, simply words that our audience have thought about and then used to describe what they’re interested in. Combine these with subsequent behaviour stats (AKA traffic stats) and we can begin to figure out content gaps and shortfalls.

Listening to your audience

Like Gerry McGovern’s Customer CareWords, relevant audience keywords also provide a pretty authentic lexicon; if you’re looking for audience relevancy, use words THEY use rather than words YOU use.

Some while ago, CDA ran some research and produced a paper called Language Pathways. While that was about aligning the relationship between SEO and brand marketing, the work left us in no doubt that we can ‘listen’ to our audiences by understanding their search terms.

This is powerful stuff.

And marrying keyword data with traffic behaviour figures gets us past the pure numbers game. It starts producing clues about what people really think about the content we produce.

A case study in keywords and content strategy

For example, Makeitandmendit.com (disclosure: I am one of the founders) currently hosts 20-25K unique visitors a month and some 60-70% of that traffic is search-generated. The analytics fascinate me and in particular the keyword logs, which I look at daily.

At first we were a little alarmed by the (relatively) high bounce rate. But looking closer, I’ve learned a few interesting things about ‘bounce’ and about how this can help structure more useful and relevant content for audiences while satisfying business imperatives.

Take ‘salt dough’ – for the uninitiated, this is a simple flour, salt and water paste that can be creatively modelled to form all sorts of decorative objects. It’s especially popular for Christmas decorations. Anyway, we produced a page on the subject in 2009 and acquired a top five Google ranking for a number of related terms. By the end of 2010, we wanted to take business advantage of this strong ranking and so we started to analyse how our content was meeting people’s expectations.

Using our GA stats for November and December 2010, we learned about (and changed) a few things:

  • Our step-by-step project instructions with pictures seemed to be pretty liked by people using focused search terms such as: ‘how to make salt dough’, ‘salt dough recipe’, ‘making salt dough’. While bounce rates were high (70-85%), we looked underneath them and saw that average time on site was around a minute and a half – about as long as it takes to read the page and even print it for reference – and 1.3 page views. This indicated that the page of content they landed on – ‘How to make salt dough’ – was pretty satisfying for this audience but not great for page views (business imperative, remember).
  • When it came to the broader terms like ‘Christmas salt dough projects’ and ‘salt dough Christmas ideas’, the bounce rate fell dramatically – around 25-35%, whacking up the page views to around 3-4 pages. Our insight was that by using this language people were probably more likely to be looking for inspiration. And so we created more projects, more pages and an inspiring ‘have a go’ kind of page – all linked, of course.
  • We’re taking this learning to other pages where we clearly have a high Google rank but are disappointing people because the content is less accessible (article style than step-by-step style). We’re also structuring content in clusters of pages rather than single pages, being careful to make each page individually valuable while helping inspire people with more general interest.

Data informs editorial strategy

There’s no rocket science here. We’ve tweaked and structured this content based on our observations. Our assessment is that now only a small percentage of people reaching our salt dough pages leave immediately. This confirms that the content we’ve produced is broadly useful to a range of different audience needs.

Combining data such as search and analytics, and looking at that information in the context of likely content engagement gave us so much more knowledge about our audiences than simply that x and y volumes arrived and left.

If we’re going to produce and manage content that people want we have to ‘listen’ to them, in their own words, to learn what they need or expect so we can respond with editorial strategies that satisfy.

Clare O’Brien (@clareob on Twitter) is a marketing and communications professional whose career has spanned several branches of the media, mostly publishing. One of her missions with CDA is to build a framework for digital content planning and measurement that clients can use to ensure their communications and services can be easily found and used, and which delivers robust budget controls.

Bloggers wanted!

We’d like to thank Clare for sharing her knowledge of analytics and content strategy with our readership of Europe-based job candidates and business clients, who look to us for practical work insights, market trends and recruitment advice in the field of content and communications.

Some of the world’s leading content and comms professionals have been kind enough to share their knowledge here, giving the inside scoop on what their work involves and the skills that go with. We’d love to keep that conversation going. So if you’re an expert in an area of tech comms, web content or have interesting things to say about working in these areas, then please do get in touch with cj@firehead.net for a guest blogger pack and a lot of encouragement.

Further reading:

 

Share Button
This entry was posted in Content Strategy, Guest Blogger, Tools and Analytics and tagged , , , , , . Bookmark the permalink.

Comments are closed.